We collected air quality data from different locations and heights around campus using DataBot and AirBeam3 sensors. The data was analyzed and visualized in MATLAB, where we also explored machine learning techniques to predict air quality patterns over time.